CN105403816A - Identification method of DC fault electric arc of photovoltaic system - Google Patents

Abstract

The invention discloses an identification method of a DC fault electric arc of a photovoltaic system. The method comprises the steps: (1) according to fundamental wave frequency and the highest harmonic time, sampling a DC current digital signal in the photovoltaic system at a sampling frequency satisfying a Shannon's sampling theorem; (2) processing acquired DC current data and then implementing a Fisher identification method; and (3) in a fault current state identification process, sampling a current, extracting a harmonic amplitude, forming a characteristic vector, and taking the characteristic vector as an input, having an identification capability, of Fisher identification method after training learning, i.e., performing an identification judgment on a fault electric arc state, and outputting a judgment result. The method is advantageous in that the Fisher identification method is used, the whole process from data input to result output does not need to set a determination threshold, the problem that the threshold is determined difficultly can be effectively prevented, and fault electric ar discrimination accuracy is improved.

Description

A kind of recognition methods of photovoltaic system DC Line Fault electric arc

Technical field

The present invention relates to a kind of recognition methods of photovoltaic system DC Line Fault electric arc, relate generally to the technical testing field of photovoltaic generating system.

Background technology

Photovoltaic generating system, as an important branch of renewable energy power generation, obtains large-scale application in different field.Along with the growth of photovoltaic generating system Applicative time, the problem of system safety operation (comprise device security problem and to operator safety problem etc.) has become a problem that can not be ignored.Wherein, the direct-current arc of photovoltaic generating system has the features such as large, the difficult extinguishing of energy, and easy initiation fire, causes system equipment and property loss.Meanwhile, because insulation breakdown causes equipment live, easily cause personal security accident.Thus to the research of photovoltaic generating system direct-current arc recognition technology, there is important theory and engineer applied value.At present, the recognition methods of photovoltaic system direct current arc fault, mostly by quantitatively calculating in time domain or frequency domain, finds out electric arc and the time and frequency domain characteristics amount that there is significant change front and back occurs, thus setting respective threshold determining whether the generation of fault electric arc.But, because photovoltaic generating system adopts solar cell as input source, the characteristic quantities such as system power, voltage, electric current affect comparatively large by factors such as illumination, temperature, environment, under different condition, photovoltaic array working point can change with external factor, and system character also changes thereupon.Therefore, judge by threshold value the method that electric arc occurs, there is threshold value and be difficult to problems such as determining, False Rate is high.

For the deficiency that current photovoltaic system DC Line Fault electric arc recognition methods exists, the invention provides a kind of without the need to threshold value setting, photovoltaic system DC Line Fault electric arc recognition methods that False Rate is low.The present invention is based on Fisher recognition methods, the harmonic component of DC current under certain condition of work is adopted to form individual features vector as characteristic quantity, by harmonic current proper vector construction feature vector space under different operating condition, respectively current characteristic vector under fault and non-faulting situation, as the input of Fisher recognition methods, is carried out pattern drill to be applied to the differentiation of photovoltaic system direct current arc fault.Different from the mode arranging threshold value in generic failure arc method for measuring, the present invention adopts the mode of sample training and study to carry out fault electric arc identification, can reduce because threshold value arranges improper and the erroneous judgement that produces and inefficacy, effectively improve the accuracy that fault electric arc detects.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of recognition methods of photovoltaic system DC Line Fault electric arc.

The present invention is achieved through the following technical solutions.

A recognition methods for photovoltaic system DC Line Fault electric arc, step comprises:

(1) DC current in photovoltaic system is sampled, if fundamental frequency f ₁with the most higher harmonics frequency n needed, then the sample frequency fs demand fulfillment of sampling to the DC current in photovoltaic system is to meet fs>2nf ₁;

(2) carrying out discrete Fourier transform (DFT) by gathering the DC current data obtained, extracting the first-harmonic I of current signal ₀and the amplitude I of front nth harmonic ₁, I ₂..., I _n, construction feature vector X=[I ₀, I ₁..., I _n], by the proper vector composition characteristic vector space that state is known, as the input of Fisher recognition methods, the state of its correspondence and fault electric arc state or non-faulting conditions at the arc are exported as recognizer simultaneously, and in this, as learning sample, complete training to Fisher recognition mechanism and study with learning sample;

(3) in fault current state recognition process, electric current is sampled, extracts harmonic amplitude and composition characteristic is vectorial, using proper vector after training study, there is the input of the Fisher recognition methods of recognition capability, namely carry out identification to fault electric arc state to judge and export to differentiate result, when detect photovoltaic generating system break down electric arc time, be then judged as fault electric arc state; When detecting that DC power-supply system does not have fault electric arc to occur, be then judged as non-faulting conditions at the arc.

Further, feature vector, X in photovoltaic DC electricity generation system by characteristic quantity I ₀, I ₁... I _nform, i.e. X=[I ₀, I ₁..., I _n], wherein n value is not less than 15.

Further, DC current is the DC current total after header box parallel connection is confluxed of photovoltaic group string branch road DC current or each branch road in (1).

Further, Fisher recognition methods in (2):

According to the state of fault current, training sample is divided into two vector space ω ₁and ω ₂, be respectively fault electric arc state and non-faulting conditions at the arc, and during the input of definition feature vector, X wherein as Fisher recognition methods, it exports y and is respectively 0 and 1.

Further, the concrete operation step of Fisher recognition methods comprises:

(1). calculate the sample average m of all kinds of training sample _iand export average y _i:

$m_i=\frac{1}{N_i}\underset{X\∈{\mathrm{\ω}}_i}{\Σ}X,y_i=\frac{1}{N_i}\underset{y\∈{\mathrm{\ω}}_i}{\Σ}y,i=1,2$

(2). calculate within-class scatter S _iwith total within-cluster variance S _w:

$S_i=\underset{X\∈{\mathrm{\ω}}_i}{\Σ}(X-m_i){(X-m_i)}^T,S_w=S_1+S_2$

(3). ask for projection vector w ^*with segmentation threshold y ₀:

$w*=S_w^{-1}(m_1-m_2)$

$y_o=\frac{N_1{y}_1+N_2{y}_2}{N_1+N_2}$

(4). after process terminates, Fisher recognition methods just has the ability identifying and judge, to the feature vector, X of any one unknown state ', first it is projected, asks for projection value y ':

y′＝w ^*X ^T

Relatively y ' and y ₀relative size can judge the multipair conditions at the arc of answering of X '.

Further, the recognition methods of above-mentioned photovoltaic system DC Line Fault electric arc is used for the DC Line Fault arc-detection of any position of photovoltaic generating system.

Beneficial effect of the present invention:

Utilize the harmonic amplitude composition characteristic vector of electric current, achieve the effective extraction to current information feature; Utilize Fisher recognition methods, from the output being input to result of data, whole process, without the need to setting decision threshold, can effectively avoid threshold value to be difficult to the problem determined, improves fault electric arc discriminant accuracy; The use of this method has generality, is trained by sample learning, can be applicable to the DC Line Fault electric arc recognition detection of any position in different photovoltaic system.

Accompanying drawing explanation

Fig. 1 is based on Fisher recognition methods photovoltaic system DC Line Fault arc-detection schematic diagram

Fig. 2 is based on Fisher method photovoltaic system DC Line Fault electric arc recognizer process flow diagram

Embodiment

According to drawings and embodiments the present invention is described in further detail below.

Step 1: the fault known to state under different condition and non-faulting DC current are sampled.

As shown in Figure 1, DC current to be measured is sampled, and carry out the pre-service such as filtering.In sampling process, according to fundamental frequency and most higher harmonics number of times, gather the DC current digital signal in photovoltaic system with the sample frequency meeting Shannon's sampling theorem, this DC current can be photovoltaic group string branch road DC current, also can be the DC current that each branch road is total after header box parallel connection is confluxed.

Step 2: utilize sampling DC current to build corresponding characteristic vector space, and carry out great amount of samples learning training as the input of Fisher algorithm, and constantly adjusting training parameter, to guarantee the accuracy of Fisher Fault Identification, builds model of cognition.

Sampling DC current data are carried out discrete Fourier transform (DFT), extracts the first-harmonic I of current signal ₀and the amplitude I of front nth harmonic ₁, I ₂..., I _n, construction feature vector X=[I ₀, I ₁..., I _n]; Then by proper vector constitutive characteristic vector space known for state, as the input of Fisher recognition methods, simultaneously using the output of the state (fault electric arc state and non-faulting conditions at the arc) of its correspondence as Fisher recognition methods, and in this, as learning sample, finally, above-mentioned learning sample is utilized to complete training to Fisher recognition mechanism and study.

Step 3: in fault current state recognition process, electric current is sampled, extracts harmonic amplitude and composition characteristic is vectorial, using proper vector after training study, there is the input of the Fisher model of cognition of recognition capability, thus identification is carried out to fault electric arc state judge and export to differentiate result.When detect photovoltaic generating system break down electric arc time, then export as fault electric arc state; When detecting that DC power-supply system does not have fault electric arc to occur, then export as non-faulting conditions at the arc, algorithm realization flow process as shown in Figure 2.

Fisher recognition methods:

According to the state of fault current, training sample is divided into two vector space ω ₁and ω ₂, be respectively fault electric arc state and non-faulting conditions at the arc, and during the input of definition feature vector, X wherein as Fisher recognition methods, it exports y and is respectively 0 and 1.

Concrete operation step comprises:

(1). calculate the sample average m of all kinds of training sample _iand export average y _i:

$m_i=\frac{1}{N_i}\underset{X\∈{\mathrm{\ω}}_i}{\Σ}X,y_i=\frac{1}{N_i}\underset{y\∈{\mathrm{\ω}}_i}{\Σ}y,i=1,2$

(2). calculate within-class scatter S _iwith total within-cluster variance S _w:

$S_i=\underset{X\∈{\mathrm{\ω}}_i}{\Σ}(X-m_i){(X-m_i)}^T,S_w=S_1+S_2$

(3). ask for projection vector w ^*with segmentation threshold y ₀:

$w*=S_w^{-1}(m_1-m_2)$

$y_o=\frac{N_1{y}_1+N_2{y}_2}{N_1+N_2}$

(4). after process terminates, Fisher recognition methods just has the ability identifying and judge, to the feature vector, X of any one unknown state ', first it is projected, asks for projection value y ':

y′＝w ^*X ^T

Above-described embodiment, only for technical conceive of the present invention and feature are described, its object is to allow the personage being familiar with this art can understand content of the present invention and be implemented, can not limit the scope of the invention with this.All equivalences done according to Spirit Essence of the present invention change or modify, and all should be encompassed in protection scope of the present invention.

Claims (6)

1. a recognition methods for photovoltaic system DC Line Fault electric arc, is characterized in that, step comprises:

(1) DC current in photovoltaic system is sampled, if fundamental frequency f ₁with the most higher harmonics frequency n needed, then the sample frequency fs demand fulfillment of sampling to the DC current in photovoltaic system is to meet fs>2nf ₁;

(2) carrying out discrete Fourier transform (DFT) by gathering the DC current data obtained, extracting the first-harmonic I of current signal ₀and the amplitude I of front nth harmonic ₁, I ₂..., I _n, construction feature vector X=[I ₀, I ₁..., I _n], by the proper vector composition characteristic vector space that state is known, as the input of Fisher recognition methods, the state of its correspondence and fault electric arc state or non-faulting conditions at the arc are exported as recognizer simultaneously, and in this, as learning sample, complete training to Fisher recognition mechanism and study with learning sample;

(3) in fault current state recognition process, electric current is sampled, extracts harmonic amplitude and composition characteristic is vectorial, using proper vector after training study, there is the input of the Fisher recognition methods of recognition capability, namely carry out identification to fault electric arc state to judge and export to differentiate result, when detect photovoltaic generating system break down electric arc time, be then judged as fault electric arc state; When detecting that DC power-supply system does not have fault electric arc to occur, be then judged as non-faulting conditions at the arc.

2. the recognition methods of photovoltaic system DC Line Fault electric arc according to claim 1, is characterized in that, feature vector, X in photovoltaic DC electricity generation system by characteristic quantity I ₀, I ₁..., I _nform, i.e. X=[I ₀, I ₁..., I _n], wherein n value is not less than 15.

3. the recognition methods of photovoltaic system DC Line Fault electric arc according to claim 1, is characterized in that, DC current is the DC current total after header box parallel connection is confluxed of photovoltaic group string branch road DC current or each branch road in (1).

4. the recognition methods of photovoltaic system DC Line Fault electric arc according to claim 1, is characterized in that, Fisher recognition methods in (2):

According to the state of fault current, training sample is divided into two vector space ω ₁and ω ₂, be respectively fault electric arc state and non-faulting conditions at the arc, and during the input of definition feature vector, X wherein as Fisher recognition methods, it exports y and is respectively 0 and 1.

5. the recognition methods of photovoltaic system DC Line Fault electric arc according to claim 4, is characterized in that, the concrete operation step of Fisher recognition methods comprises:

(1). calculate the sample average ω of all kinds of training sample _iand export average y _i:

$m_i=\frac{1}{N_i}\underset{X\∈{\mathrm{\ω}}_i}{\Σ}X,y_i=\frac{1}{N_i}\underset{y\∈{\mathrm{\ω}}_i}{\Σ}y,i=1,2$

(2). calculate within-class scatter S _iwith total within-cluster variance S _w:

$S_i=\underset{X\∈{\mathrm{\ω}}_i}{\Σ}(X-m_i){(X-m_i)}^T,S_w=S_1+S_2$

(3). ask for projection vector w ^*with segmentation threshold y ₀:

$w*=S_w^{-1}(m_1-m_2)$

$y_o=\frac{N_1{y}_1+N_2{y}_2}{N_1+N_2}$

(4). after process terminates, Fisher recognition methods just has the ability identifying and judge, to the feature vector, X of any one unknown state ', first it is projected, asks for projection value y ':

y′＝w ^*X ^T

Relatively y ' and y ₀relative size can judge the multipair conditions at the arc of answering of X '.

6. the recognition methods of the photovoltaic system DC Line Fault electric arc according to any one of claim 1-5, is characterized in that, the recognition methods of described photovoltaic system DC Line Fault electric arc is used for the DC Line Fault arc-detection of any position of photovoltaic generating system.